So you've been reviewing your eCommerce store logs and are starting to formulate a picture of shopper disengagement. Shoppers from various demographics just aren't purchasing what you thought they would be. You need to take action to increase product purchasing through online channels and suspect that they eCommerce store needs to be more personalized and targeted towards the end users needs. You're familiar with recommendation, personalization engines, but have neither the budget, nor the time to invest in building one from scratch. What with Amazon Personalize, now you don't need to. 

With Amazon Personalize, you can leverage machine learning tools to generate highly targeted and personalized product listings that help maintain end user engagement and ultimately increase product turnover. Amazon Personalize draws on the many years of knowledge and experience that Amazon has acquired running the Amazon.com eCommerce site. The high level workflow to get up and running with Amazon Personalize is this. 

Firstly, you store your inventory and demographic data in an S3 bucket. Next, use the Amazon Personalize API or JavaScript library to stream end user website activity into your Amazon Personalize endpoint. Next, you use the Amazon Personalize administration console to import your training data and specified schema. Create and train a model know as a solution. Create a campaign, which involves deploying the model created in the previous step. And then testing the campaign. To generate recommendations from the model from known users. Finally, integrate personalization and recommendations back within your website, using the get recommendations and or personalized ranking API operations. 

With Amazon Personalize, when you're configuring your solution, you can either manually select preconfigured recipe or have Amazon Personalize automatically detect and choose the right one for you via the AutoML option. A recipe consists of a recommendation algorithm and other training data required to train the solution. Having successfully integrated your recommendations into your website, continue to stream back end user activity into your solution and this allows you to fine tune and optimize the recommendations by further modeled training. The great thing about this managed service is you don't need to waste time and effort implementing infrastructure, nor setting up machine only recommendation engines or frameworks. Amazon Personalize takes care of this for you. From here on, your energies are solely focused on creating, testing and integrating recommendations and personalization into your website. In quick time, this will result in better end user engagement and growth in sales.